Evaluation of the auricular surface method for non‐adult sex estimation on the Lisbon documented collection

Abstract

ObjectivesSex is usually not estimated in skeletonized non‐adult individuals because sexual dimorphism is considered minimal before puberty. In 2017, a new approach based on the shape of the auricular surface was proposed, showing that this anatomic area of the ilium is dimorphic. This study tests the reproducibility and evaluates the accuracy of the method in a bigger sample.Materials and MethodsThirty males and 31 females from the Lisbon Collection (1 day–18 years old) were analyzed by two researchers, one experienced, and the other without previous knowledge of the method. The sample was divided into three groups according to prepubertal and pubertal hormonal peaks (<2, 0–12, and 13–18 years old). Two metric variables (DE/AD and FI/CF ratios), two morphological features (Overall Morphology, OM and Morphology of the Retroauricular End of the Superior Demiface, MRS), and two formulae were tested based on the adequate results obtained in the original article.ResultsData shows a low interobserver error (ICC > 0.92; K > 0.74). Morphological features provide better results than the metrics (DE/AD = 77.05%; FI/CF = 73.77%; OM = 80.33%; MRS = 85.24%). The discriminant function correctly classified 86.66% of the males and the logistic regression, 83.33%. The method was equally reliable in 0–12 and 13–18 years‐old age groups and more accurate in males. The accuracy of most of the variables reached 100% for males under 2 years‐old, and the probabilities were higher than for older individuals. The percentages of correct estimations are not influenced by the age, year, and cause of death.DiscussionDespite the relatively small sample size, this study confirms the usefulness of the auricular surface for non‐adult sexual estimation, especially for those younger than 2 years old. Additional validation tests in documented individuals from other geographic regions are suggested.

Evaluation of the auricular surface method for subadult sex estimation on Italian modern (19th to 20th century) identified skeletal collections

“…The current study represents the first validation study in subadults of Luna and coworkers' method (Luna et al, 2017) conducted to demonstrate the reliability of both metric and morphological variables outside of Portugal. The metric approach of Luna and co-workers (Luna et al, 2017) proved to be less effective in the Italian sample utilized in this study, partially confirming the validation study of Monge Calleja and co-workers (Monge Calleja et al, 2020). In light of our results, we suggest that the use of the metric variables may be problematic and should be cautiously applied to other skeletal populations, whereas the morphological approach yielded promising results.…”

“…Villotte (2018) applied this method on a Prehistoric Ligurian (Italy) immature individual, while Bonczarowska and co‐workers (Bonczarowska, Bonicelli, Papadomanolakis, & Kranioti, 2019) tested exclusively the morphological variables on 194 adult individuals of Greek origin from a documented modern collection, with unsatisfactory results. This method has been newly validated by Monge Calleja and co‐workers (Monge Calleja, Aranda, Santos, & Luna, 2020) on a subadult sample ranging between 0 and 18 years of age ( N = 61), from the identified skeletal collection of Lisbon (Portugal), confirming the usefulness of the auricular surface for subadult sexual estimation, although offering better results for morphological rather than metric variables. As the authors suggest additional validation tests in documented individuals from other geographic regions, the aim of the present study is to test and validate the reliability of the morphological and metric variables of this method, by analyzing a larger sample of identified individuals coming from other geographical areas.…”

“…This research attempts to apply and validate a method of subadult sex assessment based on the analysis of the auricular surface of the ilium proposed by Luna and co‐workers (2017). In the current study, the use of a larger Italian sample ( N = 127), distributed over a wider age range and a sex ratio of almost 1:1, allowed for a broad testing in a population of different geographical context than the one in which the method was implemented and validated (Luna et al, 2017; Monge Calleja et al, 2020). The reproducibility of the method is confirmed by the low interobserver errors in the Portuguese samples (Luna et al, 2017; Monge Calleja et al, 2020) and in the current research, where the replicability of the recording procedures of the features ranged from good/substantial to very high, except for inflection, for which the agreement was low.…”

“…In the current study, the use of a larger Italian sample ( N = 127), distributed over a wider age range and a sex ratio of almost 1:1, allowed for a broad testing in a population of different geographical context than the one in which the method was implemented and validated (Luna et al, 2017; Monge Calleja et al, 2020). The reproducibility of the method is confirmed by the low interobserver errors in the Portuguese samples (Luna et al, 2017; Monge Calleja et al, 2020) and in the current research, where the replicability of the recording procedures of the features ranged from good/substantial to very high, except for inflection, for which the agreement was low. This result can be attributed to difficulties in visually assessing this trait, as the irregular edges of the joint surface could be erroneously identified as presence of an inflection.…”

“…Angle vertices were finally labeled with capital letters from A to I, in order to record various measurements, using an image analysis software (Digimizer), that are used to measure seven ratios, AC/CI, FI/CF, AB/BC, HI/EH, EF/CF, DE/AD and DE/EH. As originally proposed by Luna et al (2017) and tested by Monge Calleja et al (2020), we tested a discriminant function (X = −3.470 + 0.865[DE/AD] + 2.739[FI/CF]; section point = 0.078) and a logistic regression (Psex = 1/[(1 + e‐(−25,819 + 37,723*(DE/AD) + [−2,320*(FI/CF)])]; section point = 0.5), utilizing selected measurements (DE/AD and FI/CF). Values higher that the section points indicate a male individual, while lower ones indicate female individuals.…”